Melanoma is a potentially deadly and increasingly prevalent cutaneous cancer with incidence increasing about 4% per year over the past two decades. Fortunately, melanoma is curable if detected early. The 10-year survival rate for thin melanoma is 93-99%; but for melanomas thicker than 3.6mm, it ranges from 23-68%. Unfortunately, for a non-expert, there are inherent difficulties in diagnosing lesions since early lesions can be subtle while many benign lesions may appear worrisome. This situation can pose problems for the current health care strategy under managed care in which primary care physicians (PCPs) are expected to screen suspicious pigmented lesions and refer early melanomas to dermatologists. Under this "gatekeepeer" paradigm, patients may obtain quicker access to a physician, but due to the difficulties in diagnosing melanomas, it is unclear whether their care is more cost-effective. Understanding that cost-effectiveness analyses (CEA) compare both costs and outcomes, we hypothesize that dermatologists are more cost-effective than PCPs in diagnosing and managing melanoma when considering long-term health care delivery. We assume that a relative inability to distinguish benign mimics from (pre) malignant lesions can result in unnecessary procedures. In addition, failure to detect subtle, early, surgically curable lesions may delay diagnoses of potentially high-risk lesions and result in higher costs and worse outcomes. We will address these hypotheses with decision analytic methods that will model the processes and outcomes of both health delivery strategies.